Shrinkage behavior of fly ash based geopolymer pastes with and without shrinkage reducing admixture

https://doi.org/10.1016/j.cemconcomp.2019.02.007Get rights and content

Highlights

  • Shrinkage behavior of various fly ash based geopolymer (FG) pastes was compared with that of Portland cement (PC) pastes.

  • The drying shrinkage of FG pastes increased with decreasing SiO2/Na2O mole ratios and increasing solute concentration.

  • All FG pastes displayed lower cracking potential than PC pastes.

  • Shrinkage reducing admixture (SRA) distinctly reduced shrinkage of all pastes.

  • Free drying shrinkage increases with decreasing critical pore size (ds) of the paste.

Abstract

In this study, flowability, compressive strength, drying and restrained shrinkage, cracking potential, and pore size distribution of fly ash based geopolymer (FG) pastes were investigated. The FG pastes were made with two SiO2/Na2O mole ratios (Module), two solute mass concentrations (Concentration), with and without shrinkage reducing admixture (SRA), and their properties were compared with those of Portland cement (PC) pastes. A liquid-to-binder ratio (L/B) of 0.33 was used for both PC and FG pastes. The drying condition of 25 °C and 50% RH was used for both free and restrained drying shrinkage tests. The results indicated that the Module and Concentration of activators affected all properties of FG pastes investigated. All FG pastes had a lower cracking tendency than the PC pastes, the latter of which cracked first although its shrinkage was not the highest. The drying shrinkage of FG pastes increased with decreasing Module and increasing Concentration. There was a critical pore size, ds, computed from both the water loss and pore size distributions, which was strongly related with free drying shrinkage of all the pastes studied. The smaller the ds value, the larger the free drying shrinkage a paste will have. SRA substantially coarsened pore structure, significantly reduced free drying shrinkage, noticeably reduced restrained ring shrinkage, distinctly delayed crack initiation time, and considerably decreased crack width of all pastes studied.

Introduction

Fly ash geopolymer (FG) produced by fly ash and alkaline activators has been extensively studied due to its environmental benefits, including lower energy requirements and lower emission of greenhouse gases, when compared with the manufacturing of Portland cement. Freshly cast Portland cement concrete shrinks primarily due to negative capillary pressure that leads to a volume contraction of the cement paste during water loss [1,2]. When concrete is restrained, tensile stresses develop and can lead to cracking of the concrete. Therefore, shrinkage behavior becomes a critical concern for those charged with designing durable concrete.

FG can be used as an alternative binder to Portland cement (PC) in concrete. As applications of FG increase, the shrinkage behavior of FG concrete has drawn greater attention. Several studies indicated that FG showed only slightly lower drying shrinkage than PC pastes due to the geopolymer matrix's less well-connected capillary network [3,4]. Lower shrinkage is believed to be more likely in geopolymer with a low activator/binder ratio, low sodium silicate (waterglass) modules, and low activator concentration [5,6].

Shrinkage of geopolymer occurs not only due to water loss over time during reaction and evaporation, but also because the pore structure depends on several critical factors such as alkaline activator, water content, binder materials, and curing condition [4]. The pores in geopolymer are generated during reaction, mainly from the small disordered zeolites, which makes it fundamental to understand the correlation between mix design parameters and pore structure as part of the mechanism of geopolymer shrinkage behavior.

A few researchers have studied the effects of the SiO2/Na2O mole module (Module) and concentration of solute in alkaline solution (Concentration) on FG shrinkage and corresponding cracking behavior. Puertas et al. [7] investigated the influences of fibers on drying shrinkage of geopolymer, and they reported that polypropylene fibers slightly decreased shrinkage in water glass-activated mortars cured at 50% relative humidity (RH). Bakharev et al. [8] studied the effects of different admixtures on shrinkage in water glass activated slag concrete, and they concluded that gypsum reduces both autogenous and drying shrinkage attributed to the formation of expansive phases, such as ettringite. The addition of shrinkage reducing admixture (SRA) in Portland cement systems can decrease capillary stress in pore water, which in turn reduces shrinkage and shrinkage cracking of cement based materials [9,10]. Vašíčková [11] studied the efficiency of poly (propylene glycol) based SRA and the influences of different alkali activator types and the silicate module on drying shrinkage behavior of alkali activated blast furnace slag (BFS) mortars. The author found that the high amount of alkalis had a positive influence on the effectiveness of SRA. However, the effects of SRA on shrinkage of FG have not been widely reported, hence the focus of this study on FG pastes with and without SRA.

Some work has been done on the drying shrinkage of FG [12,13], but very few results have been published on the restrained shrinkage of FG or on the addition of SRA. The effectiveness of SRA in reducing early age shrinkage should be evaluated based on both free drying and restrained drying shrinkage tests. This is because reduction in free drying shrinkage does not necessarily indicate an overall reduction in crack tendency, while restrained shrinkage tests, involving not only the shrinkage but also the strength and creep behaviors of the tested materials, provide a better indication of the material crack tendency. In addition, it is well-known that water movement is a primary cause of cement paste shrinkage [14]. However, at a given environmental condition, the amount of water moved from a cement paste is largely depending upon the characteristics of the paste pore system, e. g., the pore size distribution [5,15]. This paper attempted to provide insights on the effects of activators (i.e., Module, Concentration, and SRA) on flowability, drying shrinkage, cracking potential, and pore size distribution of FG pastes by comparing them with PC pastes as a reference for further application of FG. The mechanism of shrinkage behavior for FG was considered as well.

Section snippets

Materials and proportions

Class C fly ash, complying with ASTM C618 [16], and Type I/II Portland cement, complying with ASTM C150 [17], were used as the geopolymer and a reference binder material, respectively, for this study. The chemical compositions of the fly ash and cement are shown in Table 1. The fly ash had a specific gravity of 2.52 and a fineness of 419.6 m2/kg, and the cement had a specific gravity of 3.15 and a fineness of 429 m2/kg.

The activators used consisted of Na2SiO3 and NaOH, with SiO2/Na2O mole

Flowability

Fig. 2 reports the average slump flow of each mixture studied. None of the pastes exhibited noticeable bleeding during the mini-slump test. The flowability of the geopolymer pastes varied with Concentration, Module, and addition of SRA. As seen in the figure, the activator Concentration influenced the paste flowability more significantly than Module. When compared with PC pastes, the pastes with 25% Concentration significantly reduced slump flow, while the pastes with 20% Concentration

Conclusions

In this study, flowability, compressive strength, drying and restrained shrinkage, cracking potential, and pore size distribution of FG pastes were investigated. The FG pastes were made with two different SiO2/Na2O mole ratios (Module), solute mass concentrations (Concentration), with and without SRA, and their properties were compared with those of PC pastes. The L/B of both PC and FG pastes was 0.33. These conclusions were drawn from the investigation:

  • (1)

    The Module and Concentration of

Acknowledgments

This is a part of a dissertation of the first author, Dr. Y. Ling. Sponsorship from the Department of Civil, Construction, and Environmental Engineering at Iowa State University (ISU) for his PhD study is gratefully acknowledged. This study was started while the third author, Dr. C. Fu, was visiting ISU, and he would like to acknowledge Zhejiang University of Technology, China, for their sponsorship of his visit. Special thanks are given to Mr. Robert F. Steffes and Dr. Gilson R. Lomboy for

References (32)

  • P. Soroushian et al.

    Control of plastic shrinkage cracking with specialty cellulose fibers

    ACI Mater. J.

    (1998)
  • A. Castel et al.

    Creep and drying shrinkage of A blended slag and low calcium fly ash geopolymer concrete

    Mater. Struct.

    (2016)
  • F. Collins et al.

    Effect of pore size distribution on drying shrinking of alkali-activated slag concrete

    Cement Concr. Res.

    (2000)
  • P. Puertas et al.

    Behaviour of alkaline cement mortars reinforced with acrylic and polypropylene fibers

    Mater. Constr.

    (2000)
  • N.S. Berke et al.

    New developments in shrinkage- reducing admixtures

  • J. Mora et al.

    The influence of shrinkage-reducing admixtures on P lastic shrinkage

    Mater. Constr.

    (2003)
  • Cited by (81)

    View all citing articles on Scopus
    View full text